Informally, the required samples are built up from the known haplotypes in HapMap. Consider first the simulation of control samples in a region of the genome. A particular control individual is simulated by separately simulating its two haplotypes in the region. Each of these haplotypes is made up as mosaics of the known haplotypes in HapMap, with the mechanism for constructing these mosaic haplotypes based on population genetics theory. Fine-scale estimates of recombination rates are used to calculate the probability of breaks in the mosaic pattern as one moves along the region.
